Real Data Mining project

[supadhyaya]

I am currently working on real data set to find patterns of online workers.

Scenario : Information of workers including the past jobs done , their time taken , skill values , etc are there in a DW.

Problem : Where to start off to find the patterns of workers who regularly fail to perform work accurately. Preliminary task is to score each worker as efficient or not based upon their past history of works done.

Tables :

Worker.worker_id (int) , Worker.job_id (int) , Worker.Skill_value(quantitative value of worker's skill) (int) ,Worker.duration (time) , Worker.accuracy (int).

Important Note : Worker.accuracy has values ranging from 0-1 for all workers based on previous work done.

Anyone who is kind enough to guide this project , please let me know.

Thanking you in advance
Sanjiv U.

[editor]

this seems like a classification task.
Label workers according to whether they performed work accurately,
then use classification tools like a decision tree or regression - see
http://www.kdnuggets.com/software/classification.html